Too Many MCP Tools? How to Fix MCP Token Bloat
Too many MCP tools can burn tens of thousands of tokens before your first prompt. Audit your MCP token usage and fix context bloat with four steps.
localskills.sh Blog9 min read
ai-coding
54 stories
How to Build an MCP Server in TypeScript, Step by Step
Learn how to build an MCP server in TypeScript: define tools with Zod schemas, add resources and prompts, test with MCP Inspector, and deploy over Streamable HTTP.
localskills.sh Blog9 min read
MCP Transports: stdio vs Streamable HTTP vs SSE
MCP transports compared: stdio for local servers, Streamable HTTP for remote, and what the SSE deprecation means — with a decision tree and migration steps.
localskills.sh Blog9 min read
Private Skill Registry: Host Internal Agent Skills
Set up a private skill registry for your team's internal agent skills: compare git, self-hosted, and hosted options with SSO, roles, and audit logs.
localskills.sh Blog8 min read
How to Share Claude Skills With Your Team
Learn how to share Claude skills with your team: git repos, plugin marketplaces, and hosted registries, plus versioning, rollout, and usage analytics.
localskills.sh Blog8 min read
How to Build a Private Claude Code Plugin Marketplace
Build a private Claude Code plugin marketplace with marketplace.json in a git repo: setup, authentication, versioning, and a hosted alternative.
localskills.sh Blog8 min read
How to Share MCP Servers Across Your Engineering Team
Stop MCP config drift. Learn how to share MCP servers across your team with version-controlled .mcp.json, safe secrets, and server allowlists.
localskills.sh Blog8 min read
How to Debug an MCP Server with MCP Inspector
Debug an MCP server fast with MCP Inspector: launch it against stdio and remote servers, test tools interactively, read MCP logs, and catch failures in CI.
localskills.sh Blog10 min read
CodeHealth as a Prerequisite and Compass for Coding Agents
Healthy code works better for both agents and humans 1. Healthy code is more AI-friendly: it is easier for coding agents to analyze, modify, and extend without introducing unintended side effects. High-quality code has never been more important than in the AI era.2.
CodeHealth MCP Server by CodeScene Blog5 min read
Announcement: Deterministic PR Refactoring Agents
CodeScene’s new PR Refactoring Agent analyzes the branch using deterministic Code Health data, performs targeted refactorings, and commits improvements back to the PR. The result is code that’s easier to review, AI-friendly, and more token efficient.
CodeHealth MCP Server by CodeScene Blog4 min read
Unhealthy Code is Burning Your Token Usage - Here's the Data
Our research across C++, Java, and Python shows that agents working on unhealthy codebases consume 35-45% more tokens to complete the same tasks. That’s in addition to the increased defect risk covered in earlier research. You're not just getting worse output.
CodeHealth MCP Server by CodeScene Blog5 min read
Managing Several Coding Agents at Once
Give coding agents separate workspaces, keep track of blocked tasks, and review changes while other agents keep working.
Superset Blog3 min read
Working with Git Worktrees in Superset
Create separate workspaces for coding agents, review their diffs, and merge changes. Learn what Git worktrees isolate and which resources remain shared.
Superset Blog3 min read
Our Plan for Running 100 Coding Agents
Right now at Superset, we're able to reliably manage 5-7 coding agents in parallel - whether that's Claude Code, Codex, etc. - at a time. Our goal is to be able to manage 100 coding agents in parallel each by the end of 2026.
Superset Blog7 min read
How to get hit (probably in the face but anywhere else works too)
I ran out of technical content to write for the blog so here is something I am interested in instead.
Superset Blog2 min read
Git Worktrees: The Feature That Waited a Decade for Its Moment
Git worktrees have been around since 2015. For most of that time, they were a curiosity. Now they're having a moment. This is the story of how they came to be and why they suddenly matter.
Superset Blog8 min read
Keeping Terminal Sessions Alive Across Electron App Restarts
How we built a process-isolated terminal host that survives app restarts, handles backpressure gracefully, and enables cold restore from disk.
Superset Blog5 min read
Trae Cue Major Update
Trae Cue gets its biggest monthly update with major improvements to model performance and speed.
Trae Blog4 min read
Showing 54 of 54 stories
You're all caught up.
AiByts Weekly Brief
Get the top AI news and tools straight to your inbox every week.
Weekly · AI news
Join 15,000+ AI enthusiasts